Electronic device for supporting artificial intelligence agent services to talk to users

Abstract

An electronic device and method are provided. The method includes identifying a speech section of a user and a speech section of a neighbor in a received audio signal, identifying a user utterance in the speech section of the user and a neighbor answer to the user utterance in the speech section of the neighbor, obtaining preference information associated with the user utterance, giving a first reliability to the neighbor answer and a second reliability to an agent answer of an artificial intelligence agent generated in response to the user utterance, based on the preference information, not responding to the user utterance when the second reliability is lower than the first reliability, and outputting the agent answer when the second reliability is equal to or higher than the first reliability.

What is claimed is: 
     
       1. An electronic device comprising:
 a speaker; 
 a microphone; 
 an audio connector; 
 a wireless communication circuit; 
 a processor configured to be operatively connected to the speaker, the microphone, the audio connector, and the wireless communication circuit; and 
 a memory configured to be operatively connected to the processor, wherein the memory stores instructions that, when executed, cause the processor to:
 identify a speech section of a user and a speech section of a neighbor in an audio signal received through the microphone, the audio connector, or the wireless communication circuit, 
 identify a user utterance in the speech section of the user and a neighbor answer to the user utterance in the speech section of the neighbor, 
 obtain user preference information associated with the user utterance, 
 determine a first reliability value to the neighbor answer and a second reliability value to an agent answer of an artificial intelligence (AI) agent generated in response to the user utterance, based on the user preference information, wherein the first reliability value and the second reliability value are numerical values determined by a reliability measurement method, 
 compare the first reliability value with the second reliability value; and 
 not respond to the user utterance when the second reliability value is lower than the first reliability value, and 
 output the agent answer through the speaker, the audio connector, or the wireless communication circuit when the second reliability value is equal to or higher than the first reliability value. 
 
 
     
     
       2. The electronic device of  claim 1 , wherein the instructions cause the processor to obtain the user preference information associated with the user utterance using an artificial intelligence model personalized in relation to a preference of the user. 
     
     
       3. The electronic device of  claim 2 , wherein the instructions cause the processor to obtain the user preference information associated with the user utterance using an artificial intelligence model generalized in relation to a preference of a plurality of unspecified persons when there is no artificial intelligence model personalized to the user. 
     
     
       4. The electronic device of  claim 2 , wherein the instructions cause the processor to:
 identify a positive or negative response of the user to the output agent answer in the speech section of the user, and 
 update the personalized model, based on the identified response. 
 
     
     
       5. The electronic device of  claim 1 , wherein the instructions cause the processor to:
 configure the AI agent in a conversation mode of participating in a conversation between the user and the neighbor when a designated first utterance is identified in the speech section of the user, and 
 terminate the conversation mode when a designated second utterance is identified in the speech section of the user. 
 
     
     
       6. The electronic device of  claim 5 , wherein the instructions cause the processor to output a designated agent answer through the speaker, the audio connector, or the wireless communication circuit when a designated third utterance is identified in the speech section of the user while the AI agent is configured in the conversation mode. 
     
     
       7. The electronic device of  claim 1 , wherein the instructions cause the processor to:
 output the user utterance and the neighbor answer through a display, and 
 output the agent answer through the display when the second reliability is equal to or higher than the first reliability. 
 
     
     
       8. The electronic device of  claim 1 , wherein the instructions cause the processor to identify the speech section of the user and the speech section of the neighbor in the audio signal using an artificial intelligence model trained to find a voice of the user. 
     
     
       9. A method for operating an electronic device, the method comprising:
 identifying a speech section of a user and a speech section of a neighbor in an audio signal received through a microphone, an audio connector, or a wireless communication circuit provided in the electronic device; 
 identifying a user utterance in the speech section of the user and a neighbor answer to the user utterance in the speech section of the neighbor; 
 obtaining user preference information associated with the user utterance; 
 determining a first reliability value to the neighbor answer and a second reliability value to an agent answer of an artificial intelligence (AI) agent generated in response to the user utterance, based on the user preference information, wherein the first reliability value and the second reliability value are numerical values determined by a reliability measurement method; 
 comparing the first reliability value with the second reliability value; and 
 outputting the agent answer through a speaker, the audio connector, or the wireless communication circuit when the second reliability value is equal to or higher than the first reliability value, without responding to the user utterance when the second reliability value is lower than the first reliability value. 
 
     
     
       10. The method of  claim 9 , wherein the obtaining of the user preference information comprises obtaining the user preference information associated with the user utterance using an artificial intelligence model personalized in relation to a preference of the user. 
     
     
       11. The method of  claim 10 , further comprising:
 identifying a positive or negative response of the user to the output agent answer in the speech section of the user; and 
 updating the personalized model, based on the identified response. 
 
     
     
       12. The method of  claim 11 , wherein the identifying of the positive or negative response comprises recognizing a voice signal indicating a user emotion. 
     
     
       13. The method of  claim 10 , further comprising generating a second answer candidate of the AI agent based on the personalized AI model. 
     
     
       14. The method of  claim 9 , further comprising:
 configuring the AI agent in a conversation mode of participating in a conversation between the user and the neighbor when a designated first utterance is identified in the speech section of the user; and 
 terminating the conversation mode when a designated second utterance is identified in the speech section of the user. 
 
     
     
       15. The method of  claim 9 , further comprising:
 outputting the user utterance and the neighbor answer through a display; and 
 outputting the agent answer through the display when the second reliability is equal to or higher than the first reliability. 
 
     
     
       16. The method of  claim 9 , wherein the identifying of the speech section of the user and the speech section of the neighbor comprises identifying the speech section of the user and the speech section of the neighbor in the audio signal using an artificial intelligence model trained to find a voice of the user. 
     
     
       17. The method of  claim 16 , further comprising outputting the agent answer in response to no neighbor utterance being identified within a reference time from when the user utterance is identified. 
     
     
       18. The method of  claim 16 , further comprising outputting the agent answer in response to the neighbor answer being identified to include wrong information.